Wet. artikel
Aantekeningen Onderzoeksvorm: dubbelblind, double dummy, gerandomiseerd, 8 weekse multi-center studie Superiority: significant aantoen dat medicijn A beter werkt dan medicijn B Non-inferiority: significant aantonen dat medicijn A niet slechter werkt dan medicijn B : Ze wilden dus in eerste instantie aantonen dat budesonide superior was aan mesalamine. Toen bleek dat mesalamine effectiever was dan van tevoren werd gedacht ('there was a low chance proving superiority with budesonide therapie'), zijn ze geswitcht naar een non-inferior design: budesonide is niet minder effectief dan mesalamine. Safety population: alle gerandomiseerde patiënten die op z'n minst een dosis van study medicijnen had genomen, voor veiligheid, bijwerking etc 3-stage adaptive group sequential test design with 2 interim analyses: studie gedaan in 3 fases, waarbij tussen de fases door een analyse werd gedaan (de '2 interim analyses') om te kijken of ze al klaar waren of nog iets moesten aanpassen, bijv de grootte van de groep double dummy: als twee behandelingen niet identiek kunnen worden gemaakt, bijvoorbeeld omdat je bij de ene behandeling tabletten gebruikt en in de andere behandeling capsules, is het moeilijk om een onderzoek dubbel-blind te maken. de Double-Dummy techniek biedt hier een oplossing: Bij double-dummy maak je dan een placebo van zowel de capsule als het tablet en laat je mensen zowel een actief tablet als een placebo capsule of een placebo tablet en een actief capsule innemen. In deze studie kan dit betekend hebben dat de groep 1x9mg budesonide ook 2 placebo pillen gekregen om aan de 3 pillen per dag te komen Fase I/II of fase III: Aan de ene kant zeg je fase I/II, want er wordt gezegd dat de guideline zegt dat mensen met mildly active CD budesonide 9 mg/dag moeten hebben (dus een bestaande therapie), aan de andere kant zeg je fase III, want er word ook gezegd dat de standaard therapie 3x 3mg budesonide is, en dat betekent dat 9mg een 'nieuwe therapie' is ---- Fase I onderzoek Als na uitvoerig laboratoriumonderzoek vaststaat dat een nieuwe behandeling werkzaam is bij proefdieren, dan volgt de stap om deze voor het eerst bij mensen toe te passen. De onderzoekers weten dan nog niet hoe mensen erop reageren. Bij fase I onderzoek wordt bij een kleine groep patiënten nagegaan hoe zij de behandeling verdragen. Bij geneesmiddelenonderzoek kijkt men ook: *hoe het medicijn zich in het menselijk lichaam gedraagt *welke toedieningsvorm het meest geschikt is (bijvoorbeeld infuus, injectie, tablet of drankje) * Fase II onderzoek Als duidelijk is dat de nieuwe behandeling veilig is en mensen de behandeling verdragen, volgt fase II onderzoek. Het doel van fase II onderzoek is na te gaan of een nieuwe behandeling of een nieuwe combinatie van behandelingen bij een bepaalde soort kanker ook echt werkt tegen de ziekte. Vaak geeft het voorgaande onderzoek al aanwijzingen bij welke soorten kanker een behandeling mogelijk werkzaam is. Fase III onderzoek Bij fase III onderzoek wordt een standaardbehandeling vergeleken met een nieuwe behandeling of een nieuwe combinatie van behandelingen. In het ziekenhuis spreekt men ook wel van 'vergelijkend onderzoek'. Twee grote groepen patiënten met een bepaalde soort kanker worden met elkaar vergeleken. De ene groep krijgt de standaardbehandeling, de andere groep krijgt de mogelijk nieuwe behandeling. Summary Background & Aims Comparative data on budesonide vs mesalamine for the treatment of mild-to-moderately active Crohn's disease (CD) are sparse. We assessed the efficacy and safety of each therapy in patients with mildly to moderately active CD. Methods We performed a randomized, double-blind, double-dummy, 8-week, multicenter study in which 309 patients with mildly to moderately active CD received pH-modified-release oral budesonide (9 mg/day once daily or 3 mg/day 3 times daily) or Eudragit-L–coated oral mesalamine (4.5 g/day). Results The primary efficacy variable, clinical remission (defined as Crohn's Disease Activity Index ≤150), at the final visit occurred in 69.5% (107 of 154) of patients given budesonide vs 62.1% (95 of 153) of patients given mesalamine (difference, 7.4%; 95% repeated confidence interval, −4.6% to 18.0%; P = .001 for noninferiority). Clinical remission rates did not differ significantly between the 2 budesonide groups. Treatment response, defined as Crohn's Disease Activity Index of 150 or less and/or a decrease of 70 or more (Δ70) or 100 or more (Δ100) points from baseline to final visit, did not differ significantly between patients given budesonide vs mesalamine (Δ70, P = .11; Δ100, P = .15), or between the 2 budesonide groups (Δ70, P = .38; Δ100, P = .78). No other efficacy end points differed significantly between groups. Discontinuation because of adverse events occurred in 3% and 5% of budesonide- and mesalamine-treated patients, respectively. There were no clinically relevant differences in adverse events between the 2 budesonide groups. Conclusions Budesonide (9 mg/day) was numerically, but not statistically, more effective than Eudragit-L–coated mesalamine (4.5 g/day) in patients with mildly to moderately active CD. Budesonide (9 mg/day), administered once daily, was as effective as the standard (3 times daily) regimen. Abbreviations used in this paper *CI, confidence interval; *CRP, C-reactive protein; *CYP, cytochrome P450; *ITT, intent-to-treat; *LOCF, last observation carried forward; *MDR1, multidrug resistance 1 gene; *OD, once daily; *PP, per protocol See related article, Jensen MD et al, on page 124 in CGH. Background Uncertainty about the etiology of Crohn's disease (CD) means that treatment decisions are made empirically. Selecting the appropriate regimen for an individual patient, however, can be complex because it needs to take into account the activity, localization, and behavior of the disease; the balance between drug potency and adverse events; previous response to treatment; and the presence of extraintestinal manifestations or complications. Recently, guidelines from the European Crohn's and Colitis Organisation1 recommended that mildly active localized ileocecal CD should be treated with budesonide 9 mg/day based on evidence that it offers superior efficacy to placebo and mesalamine.2 For moderately active, localized ileocecal disease, the European Crohn's and Colitis Organisation guidelines recommend treatment with either budesonide 9 mg/day or systemic corticosteroids. Budesonide has a superior side-effect profile to conventional steroid therapy2because of its relatively low bioavailability,3 and is better able to preserve adrenal function2 and bone mass.4Superior tolerability might be attributed to extensive first-pass metabolism of budesonide by cytochrome P450 3A (CYP3A) enzymes5 and to gastrointestinal efflux mediated by P-glycoprotein, a product of the multidrug resistance 1 gene (MDR1).6 Mesalamine is regarded as showing only limited value in mild-to-moderately active CD on the basis of a meta-analysis of 3 studies conducted in patients with active ileal or colonic CD that concluded that ethylcellulose-coated mesalamine 4 g/day was associated with only a marginal benefit compared with placebo.7 However, only a single randomized study published by Thomsen et al8 a decade ago has directly compared the efficacy and safety of budesonide vs mesalamine for the management of active CD. In that trial, 182 patients with active CD (Crohn's Disease Activity Index CDAI, 200–400) received either a controlled ileal-release budesonide formulation 9 mg once daily (OD) or 2 g ethylcellulose-coated mesalamine twice daily. After 8 weeks of treatment, clinical remission was observed significantly more frequently with budesonide than mesalamine (69% vs 45%; P = .001), a difference that was sustained at week 16 (62% vs 36%; P < .001). Since then, no comparative studies of budesonide vs mesalamine have explored the use of alternative formulations to those used in the Thomsen et al8 study or examined different budesonide dosing regimens. We report here the findings of a double-blind, double-dummy study in which patients with mildly to moderately active CD were randomized to pH-modified-release budesonide (9 mg/day, given in a single dose or 3 times daily) or Eudragit-L–coated (Evonik, Essen, Germany) oral mesalamine tablets at a dose of 4.5 g/day. The primary objective was to assess the efficacy and safety of each regimen during an 8-week treatment period. ---- Materials and Methods Study Design and Conduct This was a double-blind, double-dummy, randomized, active-controlled, 8-week, phase III study conducted during the period from November 2004 to May 2008 at 46 gastroenterology centers in 7 countries (Croatia, Czech Republic, Germany, Greece, Hungary, Israel, Slovak Republic). In this 3-arm trial, patients were randomized (1:1:2 ratio) to receive budesonide (Budenofalk 3-mg capsules; Dr. Falk Pharma GmbH, Freiburg, Germany) at an oral dose of either 3 × 3 mg/day or 1 × 9 mg/day, or oral mesalamine (Salofalk 500-mg tablets; Dr. Falk Pharma GmbH) 3 × 1.5 g/day. A randomization list was generated by computer using blocks of 4 (RANCODE 3.6; IDV, Gauting, Germany) and was used to dispense study medication to the investigating center according to each patient's unique randomization number. The study was conducted in accordance with the principles of Good Clinical Practice, the Declaration of Helsinki, and all applicable national laws, after approval by independent ethics committees at each of the participating centers. Written informed consent was obtained from all patients. The EudraCT number is 2004-001213-34 (https://eudract.emea.europa.eu/) and the ClinicalTrials.gov identifier is NCT00300118(http://clinicaltrials.gov/). Patients Patients aged 18–70 years were eligible to take part in the study if they had experienced symptoms of CD for at least the preceding 3 months. Diagnosis was to be confirmed by endoscopic and histologic criteria, or by endoscopic and radiologic criteria, with localization of CD in the terminal ileum and/or ascending colon or distal colon. Patients were to be in the active phase of the disease, defined as CDAI score greater than 200 and less than 400. The main exclusion criteria were recognized CD lesions in the upper gastrointestinal tract with current symptoms; CD in the rectum; short-bowel syndrome; abscess, perforation, or active fistulas; ileostomy or colostomy; resection of more than 50 cm of the ileum; clinical signs of stricturing disease; suspected ileus; abnormal renal or hepatic function; treatment with immunosuppressants, cytostatics, methotrexate, or cyclosporine within the past 3 months (in case of treatment with azathioprine or 6-mercaptopurine, the drugs were to have been used only for maintenance of remission and dosage was to be unchanged within the preceding 3 months and during the study); anti–tumor necrosis factor-α therapy within the past 6 months; patients known to be steroid-refractory or steroid-dependent; use of conventional steroid therapy within the preceding 2 weeks; or oral budesonide greater than 6 mg/day or oral mesalamine greater than 3 g/day within the preceding 2 weeks. Evaluation Study visits took place at baseline (week 0) and weeks 2, 4, and 6, with a final visit at week 8. Each visit included assessment of CDAI, laboratory parameters, vital signs, adverse events, concomitant therapy, and compliance. Patients were issued with a diary to be completed each day. Compliance was assessed by pill counts and by daily study drug intake as recorded in the patient diaries. CDAI score was calculated retrospectively based on diary entries for the preceding 7 days. Measurement of erythrocyte sedimentation rate and urinalysis was undertaken locally, all other blood analyses were performed centrally. Blood samples for morning serum cortisol level had to be taken between 7:00 and 10:30 am, and cortisol concentration was evaluated only for patients sampled within this time window. The normal reference range for morning serum cortisol is 5–25 μg/dL.9 At the final visit, physicians completed the Physician's Global Assessment questionnaire. The primary efficacy variable was clinical remission, defined as CDAI of 150 or less, at the week-8 or withdrawal visit. Prespecified subgroup analyses were performed for the primary efficacy variable to assess the consistency of treatment effects across various patient populations. Secondary end points were treatment response, time to clinical remission/treatment response; change in CDAI score, and therapeutic success or therapeutic benefit on the Physician's Global Assessment scale. The time to remission or response was calculated as the interval from baseline to the first recording of a CDAI score of 150 or less, or a decrease of 70 or more (Δ70) or 100 or more (Δ100) points, based on daily diary entries. Safety end points included adverse events (incidence, type, and severity) and laboratory parameters. At baseline, subjects were genotyped for MDR1 3435C>T, and 2677G>T/A.10MDR1 genotypes were stratified according to haplotypes predicting low or high intestinal expression of P-glycoprotein and function.11 For genetic variants in CYP3A, subjects were genotyped for CYP3A4*1B (−392A>G) and forCYP3A5*3 (intron 3 A>G).12CYP3A5 genotypes were grouped into expressors (CYP3A5*1/*3) and nonexpressors (CYP3A5*3/*3). All genotypes were in Hardy–Weinberg equilibrium. Statistics The primary aim of the study was to show that the rate of clinical remission at the final visit was superior with budesonide (both dose groups combined) vs mesalamine. Based on data in the literature, the sample size calculation assumed that the remission rate would be 55%–60% in budesonide-treated patients and 40%–45% in the mesalamine group (ie, there would be a clinically relevant difference of 15% in the remission rate between treatments). The sample size calculation yielded a total of 368 patients, allowing for drop-outs and an α value of .025 (1-sided).13 The trial was conducted using a 3-stage adaptive group sequential test design with 2 interim analyses that could lead either to adaptation of the sample size or to cessation of the trial because of successful rejection of the null hypothesis.14 In addition to the adjusted P values, repeated confidence intervals (CIs)15 (normal approximation) were calculated to allow an estimation of the treatment effect. The option to switch to a noninferiority analysis if the remission rate with mesalamine was considerably higher than expected, in which case there would be a low chance of proving superiority with budesonide therapy, was prespecified in the study protocol. The crucial requirements for the switch from superiority to noninferiority were fulfilled.16 The trial was properly designed and performed in accordance with the requirements of a noninferiority trial. The noninferiority margin of -10% was predefined in the protocol. This was considered a reasonable margin given an assumed remission rate with placebo and mesalamine of 18%–33% [17] and [18] and 40%–45%, [18] and [19] respectively. Statistical significance was demanded for both the intention-to-treat and the per-protocol analysis. Adequate (assay) sensitivity (ie, the ability to detect relevant differences) was ensured by the following requirement in the study protocol: the switch to noninferiority testing was allowed only if superiority of budesonide compared with mesalamine could not be shown on the 2.5% level, although the remission rate with budesonide was at least 55%. This implied that noninferiority testing could be performed only if both active treatments showed remission rates within or above the expected ranges. The required sample size based on the assumptions described earlier was regarded as adequate based on a noninferiority margin of -0.10, assuming that the remission rate with budesonide would be 5% or higher than with mesalamine. At the first planned interim analysis, involving 220 patients, the Independent Data Monitoring Committee recommended switching the testing of the primary efficacy variable from superiority to noninferiority. By using ADDPLAN (ADDPLAN GmbH, Cologne, Germany)13 they calculated the sample size required to prove noninferiority with a chance of 80% in the final analysis (overall conditional power) given the remission rates observed at the interim analysis and recommended that an additional 54 patients be enrolled for the second interim analysis and a further 54 patients for the final analysis. The study was still blinded and recruiting. Confirmatory testing of noninferiority was performed using the inverse-normal method of combining the P values of the one-sided shifted asymptotic chi-square test for comparing 2 rates and maximum likelihood estimation for the unknown parameters, according to Farrington and Manning.20 If a patient discontinued the study prematurely, the last CDAI value recorded with study medication was included (last observation carried forward LOCF method). Patients without a postbaseline CDAI value were regarded as not having shown a response to treatment. All other comparisons were of an exploratory nature. Hence, no correction of P values for multiplicity was performed. For subgroup analyses for the primary efficacy variable logistic regression was used. Comparisons of the response rates to treatment were performed using a chi-square test (2-sided) and 95% CIs. The times to remission and to treatment response were compared between groups using a generalized Wilcoxon test and the log-rank test, respectively. For the pharmacogenetic assessment, the results in all patients on budesonide (3 mg 3 times a day or 9 mg OD) were analyzed. Hardy–Weinberg equilibrium was tested using de Finetti program (http://ihg.gsf.de/cgi-bin/hw/hwa1.pl). Changes in CDAI score and C-reactive protein (CRP) level were compared between different genotypic groups using the 2-sided Wilcoxon 2-sample test. The safety population comprised all randomized patients who took at least one dose of study medication. The intent-to-treat (ITT) population comprised all patients in the safety population who had active CD at baseline, defined as CDAI of 150 or greater. The per-protocol (PP) population included all ITT patients without major protocol violations who were adequately compliant (defined as taking >75% of study medication) and provided at least one postbaseline CDAI value with study medication. Results Patient Population In total, 311 patients were recruited (Figure 1). Two patients randomized to budesonide 3 mg, 3 times a day, took no verum study medication and were not evaluable. The remaining 309 patients formed the safety population (79 budesonide 3 mg 3 times a day, 77 budesonide 9 mg OD, and 153 mesalamine). Two patients randomized to budesonide had baseline CDAI of less than 150 and were excluded from the ITT population (78 budesonide 3 mg, 3 times a day, 76 budesonide 9 mg OD, and 153 mesalamine). Twelve patients in the budesonide 3 mg 3 times a day group, 8 patients in the budesonide 9 mg OD group, and 34 patients in the mesalamine group had relevant protocol violations, most frequently use of prohibited concomitant medication and premature discontinuation of the study, such that the PP population included 253 patients (Figure 1). Fifty-five patients discontinued the trial prematurely (12 budesonide 3 mg, 3 times a day; 9 budesonide 9 mg OD, and 34 mesalamine 1.5 g 3 times a day), with 86.5% and 77.8% of budesonide and mesalamine patients completing the study, respectively (Figure 1). ;http://www.sciencedirect.com/science/article/pii/S001650851001601X#gr1 : Figure 1. Patient disposition. View high quality image (186K) Demographics and baseline characteristics of the treatment groups were rather similar (Table 1). Ten patients received concomitant azathioprine or 6-mercaptopurine (ITT population), with no difference between treatment groups. Overall, 79% (164 of 207) of patients with established disease received at least one treatment for acute episodes in the 2 years before baseline, comprising mesalamine, 69% (142 of 207); conventional steroids, 31% (64 of 207); budesonide, 24% (50 of 207); antibiotics, 16% (33 of 207); azathioprine, 7% (15 of 207); and other treatments, less than 6% (ITT population). Previous treatment was assessed as successful (very good, good, satisfactory) by at least 89% of the patients, and again there was no difference between treatment groups. At least one surgical procedure as a result of CD was reported for 78 patients (25%; ITT population) with the majority (n = 70) undergoing surgery more than 6 months before study entry. There were no meaningful differences in any characteristic between the 2 budesonide dosing groups (data not shown), but mean CDAI was numerically higher in the budesonide OD group at baseline (273 ± 46 vs 259 ± 44 in the 3 times a day cohort), more patients had CDAI greater than 300 (25.0% vs 16.7%), and more patients had extraintestinal manifestations (52.6% vs 42.3%). Medium CRP level (>5 to 10 mg/L) occurred more frequently in the budesonide 9 mg OD group (18.4%) than the 3 mg 3 times a day group (6.4%), but the mean of nontransformed and log-transformed values for CRP level showed no notable differences. : Table 1. Demographics and Baseline Characteristics (ITT Population) : : AZA, azathioprine; ESR, erythrocyte sedimentation rate; 6-MP, 6-mercaptopurine; SD, standard deviation. Clinical Remission Clinical remission occurred in 69.5% (107 of 154) of budesonide patients vs 62.1% (95 of 153) of mesalamine patients in the ITT population (difference, 7.4%; 95% repeated CI, −4.6% to 18.0%; P = .001 for noninferiority; delta, −10%). The corresponding rates of remission in the PP population were 72.4% (97 of 134) and 68.9% (82 of 119), respectively (difference, 3.5%; 95% repeated CI, −8.7% to 15.5%; P = .014 for noninferiority) (Figure 2). ;http://www.sciencedirect.com/science/article/pii/S001650851001601X#gr2 : Figure 2. Clinical remission at final study visit, defined as CDAI of 150 or less. RCI, repeated CI. View high quality image (71K) Clinical remission rates in the 2 budesonide dosing groups did not differ significantly: budesonide 3 mg, 3 times a day, 71.8% (56 of 78); budesonide 9 mg OD, 67.1% (51 of 76) (difference, −4.7%; 95% CI, −19.2% to 9.9%; P = .53) in the ITT population; 75.8% (50 of 66) and 69.1% (47 of 68) (difference, −6.6%; 95% CI, −21.7% to 8.4%; P = .39) in the PP population. This was observed despite the discrepancy in mean CDAI values at baseline, which was mirrored in mean values recorded at the final visit (LOCF) (budesonide 3 mg, 3 times a day, 109 ± 79; budesonide 9 mg OD, 125 ± 92). Logistic regression analyses regarding the effect of prespecified subgroups and treatment showed a trend to higher remission rates among patients with a low CDAI score (≤300) at baseline (162 of 236 68.6%) vs high CDAI score (40 of 71 56.3%; P = .087), and significantly more frequent remission in patients with a shorter disease duration (132 of 188 70.2% for <5 y vs 70 of 119 58.8% for ≥5 y; P = .030), and among those who did not undergo a surgical procedure for CD (161 of 229 70.3% vs 41 of 78 52.6% for those undergoing ≥1 procedure; P = .004) (ITT population). Women showed a higher clinical response rate with budesonide (75.0% in the ITT population, 77.8% in the PP population) than mesalamine (56.6% and 64.5%, respectively), and the interaction between sex and treatment reached significance on logistic regression analysis in the ITT population (P = .049), with a trend to significance in the PP population (P = .092). In the subgroup analysis by sex, noninferiority of budesonide was confirmed for women (ITT population, P < .001; PP population, P = .002), but not for men (ITT population, P = .171; PP population, P = .312). Of particular note, for patients with a high CDAI score (>300) or high CRP level (>10 mg/L) at baseline, the difference in the rate of clinical remission among budesonide- vs mesalamine-treated individuals was more marked than in patients with a lower level of inflammation and the differences between CDAI and CRP groups approached significance in the ITT population. A comprehensive summary of treatment effects in predefined subpopulations is given in Table 2. : Table 2. Subgroup Analysis of Clinical Remission Rates According to Characteristics at Baseline : : NOTE. Data are shown as n (%). Significant differences are shown in bold. Clinical remission was defined as CDAI of 150 or less.ESR, erythrocyte sedimentation rate. :;a :: Difference, 18.4%; 95% CI, 3.5%–33.4%; P = .0184. ;b : Difference, 26.7%; 95% CI, 0.03%–53.3%; P = .0512. ;c : Difference, 22.5%; 95% CI, 5.9%–39.2%; P = .0100. ;d : Ten patients were missing data. Secondary Efficacy End Points Treatment response, defined as CDAI of 150 or less and/or a decrease of either 70 or more or 100 or more points from baseline to the final visit, did not differ significantly between budesonide vs mesalamine (Δ70, P= .11; Δ100, P = .15), or between the 2 budesonide regimens (Δ70, P = .38; Δ100, P = .78) (Table 3). The median time to remission and treatment response also was similar between groups. The mean CDAI values decreased during the study with both budesonide and mesalamine therapy (Supplementary Figure 1), but the mean reduction from baseline to final visit (LOCF) was approximately 20 points greater in the budesonide groups (150 ± 86 in the budesonide 3 mg 3 times a day group and 148 ± 95 in the budesonide 9 mg OD group compared with 130 ± 108 points in the mesalamine group). The mean CDAI values at baseline and the final visit (LOCF) in the budesonide-treated patients were 266 ± 45 and 117 ± 86, respectively, compared with 267 ± 54 and 137 ± 104 in patients randomized to mesalamine. The difference between the budesonide-treated patients and the mesalamine-treated patients did not reach statistical significance (P = .094), whereas the mean change from baseline CDAI was highly statistically significant within all 3 groups. : Table 3. Secondary Efficacy End Points (ITT Population) : : NOTE. All between-group differences were nonsignificant. :;a :: Clinical remission (CDAI ≤150) and/or ≥70 decrease in CDAI from baseline to final visit. ;b : Clinical remission (CDAI ≤150) and/or ≥100 decrease in CDAI from baseline to final visit. ;c : Complete relief of symptoms (category 1) or marked improvement of symptoms (category 2). ;d : Complete relief of symptoms (category 1) or marked (category 2), moderate (category 3), or slight (category 4) improvement of symptoms. In terms of individual CDAI items, the number of liquid and very soft stools reduced from baseline to final visit (LOCF) in all treatment groups, but the mean improvement was more marked with budesonide treatment compared with mesalamine (Table 3). Other CDAI item scores showed no marked differences between treatment groups. Physicians rated treatment as a therapeutic success or a therapeutic benefit numerically, but not significantly, more frequently in the budesonide treatment groups vs mesalamine (Table 3). Pharmacogenetics of Budesonide Pharmacogenetic results were available for 89% (119 of 134) of all budesonide-treated patients in the PP population. Allele prevalences suggested that the distribution of this subpopulation was likely to be representative of the general population. The mean decrease in CDAI score or CRP concentration from baseline to the final visit according to MDR1 haplotypes and CYP3A4 and CYP3A5 genotypes is shown inTable 4. None of the comparisons of the decrease in CDAI or CRP between different MDR1 haplotypes (MDR1 3435TT, 2677TT vs MDR1 3435CC, 2677GG), between different CYP3A4*1B genotypes or betweenCYP3A5*3 genotypes showed a statistically significant difference. : Table 4. Decrease in CDAI Score and CRP Concentration From Baseline to Final Visit in Patients Receiving Budesonide 3 mg 3 Times/Day or 9 mg OD (PP Population) According to Different Genotypes : : NOTE. Data are shown as mean ± SD. :;a :: Low intestinal expression of P-glycoprotein. ;b : High intestinal expression of P-glycoprotein. Safety End Points Adverse events were reported in 39%, 47%, and 47% of patients in the budesonide 3 mg 3 times a day, budesonide 9 mg OD, and mesalamine groups, respectively. The corresponding incidence of suspected adverse drug reactions was 10%, 12%, and 7%. The majority of adverse events were gastrointestinal, infections, or nervous system disorders (Table 5). Two of 9 serious adverse events (both pancreatitis in the mesalamine group) were assessed as having at least a possible relationship to study drug. Apart from acne (1 patient in the budesonide 9 mg OD group), typical steroid-related adverse drug reactions such as moon face, buffalo hump, hirsutism, and striae were not observed with budesonide treatment. Discontinuation because of adverse events occurred in 3% (n = 4) and 5% (n = 8) of budesonide- and mesalamine-treated patients, respectively. : Table 5. Adverse Events by System Organ Class Occurring in 3% or More of Patients : :;a :: Spontaneous reporting by investigator; for systematic analysis of morning serum cortisol level see the Results section. For patients in whom morning serum cortisol level was measured (budesonide, n = 143; mesalamine, n = 138), the mean ± standard deviation cortisol concentration at baseline was 14.6 ± 7.2 μg/dL in the budesonide-treated patients (budesonide 3 mg 3 times a day, 13.9 ± 6.7 μg/dL; budesonide 9 mg OD, 15.5 ± 7.7 μg/dL) and 15.1 ± 6.6 μg/dL in the mesalamine group. The morning cortisol levels decreased to 9.1 ± 6.8 μg/dL with budesonide at the final visit (LOCF) but remained stable in the mesalamine group (15.9 ± 7.9 μg/dL). This was reflected in the proportion of patients with a shift from normal morning cortisol levels (defined as 5–25 μg/dL) at baseline to a below-normal level at the final visit (LOCF), which occurred in 29% of budesonide patients (31 of 106) and 3% (3 of 107) of mesalamine patients. Remarkably, the proportion of patients with a shift to below-normal morning serum cortisol level was almost identical in both budesonide arms (17 of 55 31%, budesonide 3 mg 3 times a day; 14 of 51 27%, budesonide 9 mg OD). In addition, the absolute decrease in morning serum cortisol level did not differ between budesonide 3 mg 3 times a day and 9 mg OD. No notable changes were observed in other laboratory parameters. Discussion The results of this double-blind, double-dummy, multicenter trial show that budesonide 9 mg/day (3-mg capsules) is significantly noninferior to mesalamine 4.5 g/day (500-mg tablets) for inducing remission in patients with mildly to moderately active CD within the predefined 10% noninferiority margin. The results indicate that high-dose Eudragit-L–coated mesalamine is associated with a very high rate of remission in mildly to moderately active CD (62% in the ITT population). As anticipated, budesonide 9 mg/day also was found to be highly effective, with approximately 70% of patients achieving remission within 8 weeks. Secondary efficacy end points, including physician assessments, confirmed that there was only a small and nonsignificant difference between the budesonide and mesalamine treatment groups. This unexpected finding contrasts with the results of the only previous randomized trial of budesonide vs mesalamine in this setting.8 The high remission rate observed here with budesonide is consistent with European Crohn's and Colitis Organisation recommendations that the preferred treatment for moderately active, localized ileocecal CD is budesonide 9 mg/day or systemic corticosteroids.1 We observed a particularly high response to budesonide in women, with the difference in clinical remission between the budesonide- and mesalamine-treated patients reaching significance for the women but not for the men. Whether this finding reflects a genuine treatment effect remains uncertain because other studies of budesonide for active CD have observed no sex-specific effect.[8], [21] and [22] A greater numeric advantage for budesonide in terms of clinical remission vs mesalamine was observed in patients with more severe disease, as indicated by high baseline CDAI or increased CRP or erythrocyte sedimentation rate values at baseline. This is compatible with results of the previous randomized trial of budesonide vs mesalamine published by Thomsen et al,8 which showed a significantly higher response rate for budesonide compared with mesalamine in patients with CDAI greater than 300 (budesonide, 41%; mesalamine, 11%; P = .001). Although tested only in an exploratory sense, these data suggest that once-daily or 3-times-daily administration does not affect the efficacy of budesonide. Interestingly, no efficacy end point differed significantly between the once daily and 3 times daily budesonide regimens. Indeed, the most rapid response was observed in the budesonide 9 mg OD group. Discrepancies in the patient populations at baseline (ie, higher CDAI scores and greater frequency of extraintestinal manifestations in the once-daily cohort) may have contributed to the slightly lower remission rates in the budesonide OD group. The efficacy of oral budesonide 9 mg administered once a day has been shown elsewhere using a different formulation,[17] and [22] although in previous studies this regimen was only compared with twice-daily dosing and not 3-times-daily dosing, as in the current trial. From a clinical practice perspective, it would seem justified to recommend the budesonide 9 mg OD regimen because this would be expected to improve adherence. The rate of clinical remission observed in the mesalamine cohort was higher than anticipated based on previous large, controlled studies.[8] and [18] Interpretation of this finding is not unequivocal because the study did not include a placebo arm owing to ethical considerations, but the following factors tend to indicate that this was a genuine effect of high-dose Eudragit-L–coated mesalamine: the baseline CDAI of 267 in the mesalamine group was compatible with mild to moderate disease activity and was comparable with other placebo-controlled studies.[8], [17] and [23] The mean CDAI decrease of 130 points from baseline to final visit was higher than for placebo in similar patient settings (mean CDAI decrease, 72 points17 and 45 points23), but also much more pronounced than that of ethylcellulose-coated mesalamine.[8] and [23] Our data suggest that a mesalamine-responsive phenotype of CD exists, characterized by CDAI less than 300 and low CRP level. Although this is in line with the observation that more than a quarter of patients with CD experience a mild long-term course of the disease that requires no treatment except mesalamine,24 further studies are needed to confirm our assumption. In the previous randomized study of budesonide vs mesalamine for active CD, the remission rate for mesalamine was 45% compared with the 62% seen here, whereas the remission rate for budesonide (69%) was almost identical to the current study.8 Certain differences between the trials may account for the different outcomes observed in mesalamine-treated patients. The dose in our trial, at 4.5 g/day, was slightly higher than that used in the former trial (4 g/day). Although most baseline characteristics were similar in the 2 studies, in our trial there were fewer women and patients with previous resections, a shorter duration of disease, and greater previous exposure to mesalamine at study entry. Also, our study used the Eudragit-L–coated formulation of mesalamine (Salofalk), which has a different coating system than the ethylcellulose-coated mesalamine formulation (Pentasa, Ferring, Vanløse, Denmark) administered in the previous study. The 2 galenical formulations result in dissimilar pharmacokinetic profiles: the Eudragit-L–coated formulation releases mesalamine predominantly in the terminal ileum with greater mesalamine absorption, which may be of importance in a transmural condition such as CD.25 It is noteworthy that we found no effect of putatively important CYP3A4 and CYP3A5 genotypes on the response to oral budesonide. In a clinical pharmacologic trial, intestinal CYP3A4 expression showed an impact on budesonide pharmacokinetics, whereas budesonide pharmacokinetics did not differ betweenCYP3A5 genotypic groups.26 It is well known that a 2677G>T/A single nucleotide polymorphism in exon 21 and a 3435C>T single nucleotide polymorphism in exon 26 of MDR1 affect expression of P-glycoprotein and thereby the pharmacokinetics of some commonly used drugs.27 In CD, it has been reported that a poor response to corticosteroids other than budesonide, namely prednisone and hydrocortisone, is related to increased expression of P-glycoprotein.28 In our study population, however, we observed no hint of a pharmacogenetic effect of MDR1 genetics on the response to oral budesonide therapy. The overall safety and tolerability of budesonide 9 mg/day was similar to that of mesalamine 4.5 g/day. Both drugs were well tolerated, with approximately 10% of patients experiencing adverse events with at least a possible relation to study drug and a low rate of discontinuations because of adverse events. The difference in incidence of gastrointestinal disorders between the budesonide and mesalamine groups (14% vs 24%, respectively) was owing to exacerbation and symptoms of CD that had to be reported as adverse events, and thus reflected a difference in efficacy rather than in tolerability. As expected from the pharmacodynamic profiles of the study drugs, only budesonide treatment influenced the morning serum cortisol level, with one third of budesonide-treated patients showing a decrease to below the lower limit of normal at the final visit. The observed decrease in cortisol levels is consistent with earlier findings for another oral budesonide formulation given as a regimen of 9 mg OD per day,29 which induced a markedly less pronounced suppression of pituitary-adrenal function than the conventional steroid prednisolone administered in a tapering schedule from 40 to 5 mg/d over 8 weeks. Overall, the effect of the administration schedule, OD vs 3 times a day, on the suppression of pituitary-adrenal function was similar in our study. There were no marked differences in adverse events or other safety end points between the once-daily and 3-times-daily budesonide groups. The trial used a robust double-dummy design and the treatment period of 8 weeks has been shown previously to be sufficient for achieving remission in moderately active CD.[21], [22] and [30] The use of the CDAI scale to determine clinical remission is a widely accepted approach and the CDAI-based secondary efficacy end points provided additional sensitivity for detecting a therapeutic response.31 No placebo arm was included in the trial because this was considered difficult to justify ethically in a population with high disease activity, an approach that was supported by the knowledge that study treatment would achieve remission rates of 40%–60% compared with the 20% remission rate typically seen with placebo.[18] and [21]Furthermore, the trial was planned under the primary study hypothesis that budesonide would be superior to mesalamine.8 If this hypothesis had held true, a placebo arm would have been of little extra value. The study protocol was constructed to allow conversion from the planned superiority design to a noninferiority trial, as became necessary owing to the unexpectedly high remission rates with mesalamine, and the sample size was adequate to support the final noninferiority testing. In conclusion, budesonide 9 mg/day showed numeric, but not statistical, superiority to Eudragit-L–coated mesalamine 4.5 g/day for the treatment of patients with mildly to moderately active CD. In patients with a greater degree of inflammation, there appeared to be a greater benefit for budesonide vs mesalamine, although this did not reach significance. Treatment with either agent was well tolerated, with a low rate of suspected adverse drug reactions or discontinuations owing to intolerance. Interestingly, budesonide 9 mg/day administered OD was as effective and safe as the standard 3-times-daily regimen, which may provide an opportunity to improve treatment compliance. Acknowledgments The authors give special thanks to P. Heine and R. Schwarz (medicomp GmbH, Planegg, Germany) for their assistance in conducting the clinical trial. The authors would like to thank all patients and investigators for their participation and contribution to the study. Editorial assistance, comprising revision and finalization of the draft manuscript prepared by the authors, was provided by freelance writer Caroline Dunstall, with funding from Dr. Falk Pharma GmbH.